The selection of candidates for public office is one of the pivotal functions political parties perform in liberal democracies. Many works have studied the nomination of candidates for the legislative branch. Few studies have looked into the nomination of candidates for the executive branch in presidential democracies. However, very few works have looked at candidates' nomination for the executive branch in parliamentary democracies, who have been called electoral leaders or top candidates. This dissertation contributes to filling this gap by exploring three main research questions. First, what criteria do political parties use when nominating a top candidate? Second, to what extent do political parties nominate their top candidates on electoral considerations? Third, under which conditions is electoral competition more likely to shape party decision-making? I consider that political parties hold two criteria when nominating a top candidate. First, based on top candidates' electoral and campaigning function, parties seek to nominate electable top candidates likely to achieve more votes. Second, considering that top candidates may become prime ministers after the election and perform a series of post-electoral functions, parties will seek to nominate more reliable candidates who stay close to their party’s preferences. Building on the literature on party organization change, this dissertation proposes a new theoretical framework for understanding how different incentives can drive parties to nominate top candidates closer to one or the other criterion. In particular, this dissertation studies four factors: the party’s screening and recruitment capacity, internal demand, external demand, and the type of selectorate. To test the influence of these four factors, this dissertation presents a novel dataset of more than 2500 sub-national top candidates in Canada, Germany and Spain. There have been collecting information about the personal, partisan and political background of top candidates, which been complemented with information about the type of selectorate, the party’s internal structure, and the electorate's state. The main results are summarized as follows. First, parties are heavily dependent on their access to public institutions to recruit and train new members and produce top candidates with high degrees of reliability. Second, the results show that political parties are more reactive to changes in their internal coalitions' composition than to changes in the overall electorate. Finally, the results show that party primaries tend to differ from party conferences and party elites when the party has experienced some environmental change regarding the type of selectorate. This dissertation contributes to the understanding of the role of top candidates in parliamentary democracies and academic knowledge about party organizational change and adaptation.